In [181]:
#%matplotlib inline

# import libraries 
import pandas as pd
import numpy as np
from statistics import mean

#data visualization
import matplotlib as mpl
import plotly.graph_objects as go
import plotly.express as px
import plotly.offline as pyo
import plotly.graph_objs as go
import matplotlib.pyplot as plt 
from sklearn.metrics import mean_absolute_error
# ARIMA modelling libraries
from sklearn.metrics import mean_squared_error
from math import sqrt
from statsmodels.tsa.arima.model import ARIMA

from statsmodels.tsa.stattools import acf, pacf
from statsmodels.tools.sm_exceptions import ConvergenceWarning
import warnings
import seaborn as sns
warnings.simplefilter('ignore', ConvergenceWarning)
sns.set()

import scipy
print(scipy.__version__)

from statsmodels.tsa.statespace.sarimax import SARIMAX
from pandas.plotting import autocorrelation_plot
from statsmodels.tsa.stattools import adfuller, acf, pacf,arma_order_select_ic
import statsmodels.formula.api as smf
import statsmodels.tsa.api as smt
import statsmodels.api as sm
import scipy.stats as scs
from pandas import Series as Series
from prophet import Prophet
from prophet.plot import plot_cross_validation_metric
from prophet.diagnostics import cross_validation
from prophet.diagnostics import performance_metrics

import plotly.io as pio
pio.renderers.default='notebook'
# Set notebook mode to work in offline
pyo.init_notebook_mode()

1.10.1
In [182]:
def best_fit_slope_and_intercept(xs, ys):
    m = (((mean(xs)*mean(ys)) - mean(xs*ys)) /
         ((mean(xs)*mean(xs)) - mean(xs*xs)))
    b = mean(ys) - m*mean(xs)
    
    return m, b

Temperature¶

In [183]:
# read data sets
df = pd.read_csv("Environment_Temperature_change_E_All_Data_NOFLAG.csv", encoding='latin-1')
df_countrycode=pd.read_csv('FAOSTAT_data_11-24-2020.csv')
df.head()
Out[183]:
Area Code Area Months Code Months Element Code Element Unit Y1961 Y1962 Y1963 ... Y2010 Y2011 Y2012 Y2013 Y2014 Y2015 Y2016 Y2017 Y2018 Y2019
0 2 Afghanistan 7001 January 7271 Temperature change °C 0.777 0.062 2.744 ... 3.601 1.179 -0.583 1.233 1.755 1.943 3.416 1.201 1.996 2.951
1 2 Afghanistan 7001 January 6078 Standard Deviation °C 1.950 1.950 1.950 ... 1.950 1.950 1.950 1.950 1.950 1.950 1.950 1.950 1.950 1.950
2 2 Afghanistan 7002 February 7271 Temperature change °C -1.743 2.465 3.919 ... 1.212 0.321 -3.201 1.494 -3.187 2.699 2.251 -0.323 2.705 0.086
3 2 Afghanistan 7002 February 6078 Standard Deviation °C 2.597 2.597 2.597 ... 2.597 2.597 2.597 2.597 2.597 2.597 2.597 2.597 2.597 2.597
4 2 Afghanistan 7003 March 7271 Temperature change °C 0.516 1.336 0.403 ... 3.390 0.748 -0.527 2.246 -0.076 -0.497 2.296 0.834 4.418 0.234

5 rows × 66 columns

In [184]:
#1. Renaming
df.rename(columns = {'Area':'Country Name'},inplace = True)
df.set_index('Months', inplace=True)
df.rename({'Dec\x96Jan\x96Feb': 'Winter', 'Mar\x96Apr\x96May': 'Spring', 'Jun\x96Jul\x96Aug':'Summer','Sep\x96Oct\x96Nov':'Fall'}, axis='index',inplace = True)
df.reset_index(inplace = True)
In [185]:
#2. Filtering 
df = df[df['Element'] == 'Temperature change']

#2. Drop unwanted columns from df_countrycode
df_countrycode.drop(['Country Code','M49 Code','ISO2 Code','Start Year','End Year'],axis=1,inplace=True)
df_countrycode.rename(columns = {'Country':'Country Name','ISO3 Code':'Country Code'},inplace=True)

#3. Merging with df to df_country
df = pd.merge(df, df_countrycode, how='outer', on='Country Name')

#2. Drop unwanted columns
df.drop(['Area Code','Months Code','Element Code','Element','Unit'],axis=1,inplace=True)

#3.Channing dataframe organization
df = df.melt(id_vars=["Country Code", "Country Name","Months",], var_name="year", value_name="tem_change")
df["year"] = [i.split("Y")[-1] for i in df.year]

display(df.head(5))
Country Code Country Name Months year tem_change
0 AFG Afghanistan January 1961 0.777
1 AFG Afghanistan February 1961 -1.743
2 AFG Afghanistan March 1961 0.516
3 AFG Afghanistan April 1961 -1.709
4 AFG Afghanistan May 1961 1.412
In [186]:
df0 = df[df['Country Name'] == 'World']
year_tem_change = df0.groupby(['year'])['tem_change'].mean().to_frame().reset_index()
#year_tem_change['tem_change_ma'] = year_tem_change['tem_change'].rolling(2).mean()
In [187]:
year_tem_change.head()
Out[187]:
year tem_change
0 1961 0.155941
1 1962 0.053235
2 1963 0.139000
3 1964 -0.277353
4 1965 -0.170941
In [188]:
m, b = best_fit_slope_and_intercept(pd.to_numeric(year_tem_change['year']), year_tem_change['tem_change'])

regression_line = []
for x in pd.to_numeric(year_tem_change['year']):
    regression_line.append((m*x)+b)
    
year_tem_change['line'] = regression_line
In [ ]:
In [189]:
# Create traces
fig = go.Figure()

fig.add_trace(go.Scatter(x = year_tem_change['year'] , y=year_tem_change.tem_change, mode='lines', name='Global temperature'))
fig.add_trace(go.Scatter(x = year_tem_change['year'] , y=year_tem_change.line, line = dict(color='red', width=1, dash='dash'), name='Trend'))

# adjusting size of graph, legend place, and background colour
fig.update_layout(autosize=False, width=1000, height=600, margin=dict(l=50, r=50, b=100, t=100, pad=4), 
    template='seaborn',title='Global Temperature Change (1961 - 2019)',paper_bgcolor="rgb(234, 234, 242)",
    legend=dict(orientation="h", yanchor="bottom", y=1.02, xanchor="right", x=1))
fig.update_xaxes(type='category',title='Years')
fig.update_yaxes(title='Temperature Change')

fig.show("notebook")
In [190]:
data = year_tem_change.rename(columns={'tem_change': 'value'})
data = data.astype({'year':'int'})
#data = data.drop(columns=['tem_change_ma', 'line'])
#nd_df = nd_df.iloc[60:]
#nd_df = nd_df.iloc[:-1]
data.head()
Out[190]:
year value line
0 1961 0.155941 -0.276829
1 1962 0.053235 -0.249188
2 1963 0.139000 -0.221546
3 1964 -0.277353 -0.193905
4 1965 -0.170941 -0.166264
In [191]:
data['year'] = pd.to_datetime(data['year'], format="%Y")
data.tail()
Out[191]:
year value line
54 2015-01-01 1.460647 1.215799
55 2016-01-01 1.626059 1.243441
56 2017-01-01 1.462353 1.271082
57 2018-01-01 1.278294 1.298723
58 2019-01-01 1.510882 1.326364
In [192]:
ts = data[['year', 'value']].rename(columns={'year':'ds', 'value':'y'})

ts.columns=['ds','y']
temp_model = Prophet() 
temp_model.fit(ts)

23:54:11 - cmdstanpy - INFO - Chain [1] start processing
23:54:11 - cmdstanpy - INFO - Chain [1] done processing
Out[192]:
<prophet.forecaster.Prophet at 0x150a6a3b0>
In [193]:
# predict for 1 year in the furure and MS - month start is the frequency
future = temp_model.make_future_dataframe(periods = 31, freq = 'YS') 
future.tail()
Out[193]:
ds
85 2046-01-01
86 2047-01-01
87 2048-01-01
88 2049-01-01
89 2050-01-01
In [194]:
# now lets make the forecasts
forecast = temp_model.predict(future)
forecast.tail()
Out[194]:
ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper yearly yearly_lower yearly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
85 2046-01-01 1.114602 1.863250 2.348392 1.114404 1.114803 0.991871 0.991871 0.991871 0.991871 0.991871 0.991871 0.0 0.0 0.0 2.106473
86 2047-01-01 1.142289 1.877799 2.349997 1.142081 1.142504 0.954901 0.954901 0.954901 0.954901 0.954901 0.954901 0.0 0.0 0.0 2.097190
87 2048-01-01 1.169976 1.828522 2.299027 1.169754 1.170201 0.891753 0.891753 0.891753 0.891753 0.891753 0.891753 0.0 0.0 0.0 2.061729
88 2049-01-01 1.197739 1.963256 2.442817 1.197502 1.197978 1.002622 1.002622 1.002622 1.002622 1.002622 1.002622 0.0 0.0 0.0 2.200362
89 2050-01-01 1.225427 1.970502 2.455005 1.225176 1.225680 0.991871 0.991871 0.991871 0.991871 0.991871 0.991871 0.0 0.0 0.0 2.217298
In [195]:
temp_model.plot(forecast)
Out[195]:
In [196]:
df_cv = cross_validation(temp_model, initial='730 days', period='180 days', horizon = '365 days')

  0%|          | 0/112 [00:00<?, ?it/s]
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:12 - cmdstanpy - INFO - Chain [1] done processing
23:54:12 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:13 - cmdstanpy - INFO - Chain [1] done processing
23:54:13 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:14 - cmdstanpy - INFO - Chain [1] done processing
23:54:14 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:15 - cmdstanpy - INFO - Chain [1] done processing
23:54:15 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:16 - cmdstanpy - INFO - Chain [1] done processing
23:54:16 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:17 - cmdstanpy - INFO - Chain [1] start processing
23:54:17 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:18 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:18 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:18 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:18 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:18 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:18 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:18 - cmdstanpy - INFO - Chain [1] done processing
23:54:18 - cmdstanpy - INFO - Chain [1] start processing
23:54:19 - cmdstanpy - INFO - Chain [1] done processing
23:54:19 - cmdstanpy - INFO - Chain [1] start processing
23:54:19 - cmdstanpy - INFO - Chain [1] done processing
23:54:19 - cmdstanpy - INFO - Chain [1] start processing
23:54:19 - cmdstanpy - INFO - Chain [1] done processing
23:54:19 - cmdstanpy - INFO - Chain [1] start processing
23:54:19 - cmdstanpy - INFO - Chain [1] done processing
23:54:19 - cmdstanpy - INFO - Chain [1] start processing
23:54:19 - cmdstanpy - INFO - Chain [1] done processing
23:54:19 - cmdstanpy - INFO - Chain [1] start processing
23:54:19 - cmdstanpy - INFO - Chain [1] done processing
23:54:19 - cmdstanpy - INFO - Chain [1] start processing
23:54:19 - cmdstanpy - INFO - Chain [1] done processing
23:54:19 - cmdstanpy - INFO - Chain [1] start processing
23:54:20 - cmdstanpy - INFO - Chain [1] done processing
23:54:20 - cmdstanpy - INFO - Chain [1] start processing
23:54:20 - cmdstanpy - INFO - Chain [1] done processing
23:54:20 - cmdstanpy - INFO - Chain [1] start processing
23:54:20 - cmdstanpy - INFO - Chain [1] done processing
23:54:20 - cmdstanpy - INFO - Chain [1] start processing
23:54:20 - cmdstanpy - INFO - Chain [1] done processing
23:54:20 - cmdstanpy - INFO - Chain [1] start processing
23:54:20 - cmdstanpy - INFO - Chain [1] done processing
23:54:20 - cmdstanpy - INFO - Chain [1] start processing
23:54:20 - cmdstanpy - INFO - Chain [1] done processing
23:54:20 - cmdstanpy - INFO - Chain [1] start processing
23:54:21 - cmdstanpy - INFO - Chain [1] done processing
23:54:21 - cmdstanpy - INFO - Chain [1] start processing
23:54:21 - cmdstanpy - INFO - Chain [1] done processing
23:54:21 - cmdstanpy - INFO - Chain [1] start processing
23:54:21 - cmdstanpy - INFO - Chain [1] done processing
23:54:21 - cmdstanpy - INFO - Chain [1] start processing
23:54:21 - cmdstanpy - INFO - Chain [1] done processing
23:54:21 - cmdstanpy - INFO - Chain [1] start processing
23:54:21 - cmdstanpy - INFO - Chain [1] done processing
23:54:21 - cmdstanpy - INFO - Chain [1] start processing
23:54:21 - cmdstanpy - INFO - Chain [1] done processing
23:54:21 - cmdstanpy - INFO - Chain [1] start processing
23:54:21 - cmdstanpy - INFO - Chain [1] done processing
23:54:21 - cmdstanpy - INFO - Chain [1] start processing
23:54:22 - cmdstanpy - INFO - Chain [1] done processing
23:54:22 - cmdstanpy - INFO - Chain [1] start processing
23:54:22 - cmdstanpy - INFO - Chain [1] done processing
23:54:22 - cmdstanpy - INFO - Chain [1] start processing
23:54:22 - cmdstanpy - INFO - Chain [1] done processing
23:54:22 - cmdstanpy - INFO - Chain [1] start processing
23:54:22 - cmdstanpy - INFO - Chain [1] done processing
23:54:22 - cmdstanpy - INFO - Chain [1] start processing
23:54:22 - cmdstanpy - INFO - Chain [1] done processing
In [197]:
df_cv.head()
Out[197]:
ds yhat yhat_lower yhat_upper y cutoff
0 1964-01-01 0.099364 0.046589 0.157843 -0.277353 1963-04-20
1 1964-01-01 0.099364 0.041949 0.153759 -0.277353 1963-10-17
2 1965-01-01 -0.188565 -0.320097 -0.057740 -0.170941 1964-04-14
3 1965-01-01 -0.188565 -0.324691 -0.062447 -0.170941 1964-10-11
4 1966-01-01 -0.290157 -0.414110 -0.174580 0.171412 1965-04-09
In [198]:
df_p = performance_metrics(df_cv)
df_p.head()
Out[198]:
horizon mse rmse mae mape mdape smape coverage
0 54 days 0.039432 0.198574 0.160114 1.393686 0.452779 0.545577 0.636364
1 59 days 0.054683 0.233843 0.193751 1.385720 0.452779 0.564943 0.545455
2 64 days 0.045442 0.213170 0.170163 0.366776 0.389877 0.425410 0.636364
3 70 days 0.046164 0.214859 0.177599 0.376121 0.389877 0.435598 0.636364
4 75 days 0.058282 0.241416 0.194479 0.362258 0.389877 0.410164 0.636364
In [199]:
plot_cross_validation_metric(df_cv, metric='mape')
Out[199]:
In [200]:
plot_cross_validation_metric(df_cv, metric='rmse')
Out[200]:
In [201]:
plot_cross_validation_metric(df_cv, metric='mae')
Out[201]:
In [ ]:
In [ ]:

Natural Disasters¶

In [202]:
nd = pd.read_csv('EMDAT_1900-2021_NatDis.csv')
nd.head()
Out[202]:
Dis No Year Seq Disaster Group Disaster Subgroup Disaster Type Disaster Subtype Disaster Subsubtype Event Name Entry Criteria ... End Day Total Deaths No Injured No Affected No Homeless Total Affected Reconstruction Costs ('000 US$) Insured Damages ('000 US$) Total Damages ('000 US$) CPI
0 1900-9002-CPV 1900 9002 Natural Climatological Drought Drought NaN NaN NaN ... NaN 11000.0 NaN NaN NaN NaN NaN NaN NaN 3.261389
1 1900-9001-IND 1900 9001 Natural Climatological Drought Drought NaN NaN NaN ... NaN 1250000.0 NaN NaN NaN NaN NaN NaN NaN 3.261389
2 1902-0012-GTM 1902 12 Natural Geophysical Earthquake Ground movement NaN NaN Kill ... 18.0 2000.0 NaN NaN NaN NaN NaN NaN 25000.0 3.391845
3 1902-0003-GTM 1902 3 Natural Geophysical Volcanic activity Ash fall NaN Santa Maria Kill ... 8.0 1000.0 NaN NaN NaN NaN NaN NaN NaN 3.391845
4 1902-0010-GTM 1902 10 Natural Geophysical Volcanic activity Ash fall NaN Santa Maria Kill ... 24.0 6000.0 NaN NaN NaN NaN NaN NaN NaN 3.391845

5 rows × 43 columns

In [203]:
nd_general = nd.loc[:,['Year','Disaster Subgroup','Disaster Type','Total Deaths','Total Affected']]
nd_general = nd_general[nd_general['Disaster Subgroup'] != 'Extra-terrestrial']
nd_general.head()
Out[203]:
Year Disaster Subgroup Disaster Type Total Deaths Total Affected
0 1900 Climatological Drought 11000.0 NaN
1 1900 Climatological Drought 1250000.0 NaN
2 1902 Geophysical Earthquake 2000.0 NaN
3 1902 Geophysical Volcanic activity 1000.0 NaN
4 1902 Geophysical Volcanic activity 6000.0 NaN
In [204]:
#Plot 'Number of Disasters Recorded by Year'
nd_general_count = nd_general.groupby(['Year',]).agg({'Disaster Type':'count',})
nd_general_count = nd_general_count.reset_index()
nd_general_count.head()
Out[204]:
Year Disaster Type
0 1900 7
1 1901 2
2 1902 10
3 1903 12
4 1904 4
In [205]:
fig = px.bar(nd_general_count, x="Year", y='Disaster Type',text='Disaster Type' , title="Number of Disasters Recorded by Year")
fig.update_traces(texttemplate='%{text:.2f}', textposition='outside')
# adjusting size of graph, legend place, and background colour
fig.update_layout(autosize=False, width=1000, height=600, margin=dict(l=50, r=50, b=100, t=100, pad=4),
    template='seaborn', paper_bgcolor="rgb(234, 234, 242)",
    legend=dict(orientation="v", yanchor="bottom", y=0.3, xanchor="left", x=1.02
))
In [206]:
graph_subgroups = nd.groupby(['Disaster Subgroup','Disaster Type']).size().to_frame('Number of disasters').reset_index()
graph_subgroups.sort_values('Number of disasters',ascending=False, inplace=True)
graph_subgroups.head()
Out[206]:
Disaster Subgroup Disaster Type Number of disasters
10 Hydrological Flood 5399
14 Meteorological Storm 4420
7 Geophysical Earthquake 1528
1 Biological Epidemic 1496
11 Hydrological Landslide 767
In [207]:
fig = px.bar(graph_subgroups, x="Disaster Type", y='Number of disasters',text='Number of disasters' , title="Number of Disasters Recorded by Year")
fig.update_traces(texttemplate='%{text:.2f}', textposition='outside')
# adjusting size of graph, legend place, and background colour
fig.update_layout(autosize=False, width=1000, height=600, margin=dict(l=50, r=50, b=100, t=100, pad=4),
    template='seaborn', paper_bgcolor="rgb(234, 234, 242)",
    legend=dict(orientation="v", yanchor="bottom", y=0.3, xanchor="left", x=1.02
))
In [208]:
nd_df = nd_general_count.rename(columns={'Year': 'year', 'Disaster Type': 'value'})
nd_df = nd_df.iloc[60:]
nd_df = nd_df.iloc[:-1]
nd_df.tail()
Out[208]:
year value
116 2016 349
117 2017 367
118 2018 336
119 2019 443
120 2020 397
In [209]:
nd_df['year'] = pd.to_datetime(nd_df['year'], format="%Y")

ts = nd_df[['year', 'value']].rename(columns={'year':'ds', 'value':'y'})

ts.columns=['ds','y']
nd_model = Prophet() 
nd_model.fit(ts)

23:54:23 - cmdstanpy - INFO - Chain [1] start processing
23:54:23 - cmdstanpy - INFO - Chain [1] done processing
Out[209]:
<prophet.forecaster.Prophet at 0x1368d9600>
In [210]:
# predict for 1 year in the furure and MS - month start is the frequency
future = nd_model.make_future_dataframe(periods = 30, freq = 'YS') 
future.tail()
Out[210]:
ds
86 2046-01-01
87 2047-01-01
88 2048-01-01
89 2049-01-01
90 2050-01-01
In [211]:
# now lets make the forecasts
forecast = nd_model.predict(future)
forecast.tail()
Out[211]:
ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper yearly yearly_lower yearly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
86 2046-01-01 565.568145 602.206726 755.202189 565.517439 565.626925 109.491026 109.491026 109.491026 109.491026 109.491026 109.491026 0.0 0.0 0.0 675.059171
87 2047-01-01 573.211869 597.407547 756.121482 573.157630 573.273745 107.878859 107.878859 107.878859 107.878859 107.878859 107.878859 0.0 0.0 0.0 681.090728
88 2048-01-01 580.855593 611.531398 757.500225 580.798536 580.921735 104.258946 104.258946 104.258946 104.258946 104.258946 104.258946 0.0 0.0 0.0 685.114539
89 2049-01-01 588.520259 618.307996 774.542108 588.459951 588.590299 109.093394 109.093394 109.093394 109.093394 109.093394 109.093394 0.0 0.0 0.0 697.613653
90 2050-01-01 596.163982 624.169292 787.526847 596.100177 596.237936 109.491026 109.491026 109.491026 109.491026 109.491026 109.491026 0.0 0.0 0.0 705.655009
In [212]:
nd_model.plot(forecast)
Out[212]:
In [213]:
df_cv = cross_validation(nd_model, initial='730 days', period='180 day', horizon = '365 days')

  0%|          | 0/116 [00:00<?, ?it/s]
23:54:24 - cmdstanpy - INFO - Chain [1] start processing
23:54:26 - cmdstanpy - INFO - Chain [1] done processing
23:54:26 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:28 - cmdstanpy - INFO - Chain [1] done processing
23:54:28 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:29 - cmdstanpy - INFO - Chain [1] done processing
23:54:29 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:30 - cmdstanpy - INFO - Chain [1] done processing
23:54:30 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:31 - cmdstanpy - INFO - Chain [1] done processing
23:54:31 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:32 - cmdstanpy - INFO - Chain [1] done processing
23:54:32 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:33 - cmdstanpy - INFO - Chain [1] done processing
23:54:33 - cmdstanpy - INFO - Chain [1] start processing
23:54:34 - cmdstanpy - INFO - Chain [1] done processing
23:54:34 - cmdstanpy - INFO - Chain [1] start processing
23:54:34 - cmdstanpy - INFO - Chain [1] done processing
23:54:34 - cmdstanpy - INFO - Chain [1] start processing
23:54:34 - cmdstanpy - INFO - Chain [1] done processing
23:54:34 - cmdstanpy - INFO - Chain [1] start processing
23:54:34 - cmdstanpy - INFO - Chain [1] done processing
23:54:34 - cmdstanpy - INFO - Chain [1] start processing
23:54:34 - cmdstanpy - INFO - Chain [1] done processing
23:54:34 - cmdstanpy - INFO - Chain [1] start processing
23:54:34 - cmdstanpy - INFO - Chain [1] done processing
23:54:34 - cmdstanpy - INFO - Chain [1] start processing
23:54:34 - cmdstanpy - INFO - Chain [1] done processing
23:54:34 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:35 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:35 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:35 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:35 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:35 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:35 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:35 - cmdstanpy - INFO - Chain [1] start processing
23:54:35 - cmdstanpy - INFO - Chain [1] done processing
23:54:36 - cmdstanpy - INFO - Chain [1] start processing
23:54:36 - cmdstanpy - INFO - Chain [1] done processing
23:54:36 - cmdstanpy - INFO - Chain [1] start processing
23:54:36 - cmdstanpy - INFO - Chain [1] done processing
23:54:36 - cmdstanpy - INFO - Chain [1] start processing
23:54:36 - cmdstanpy - INFO - Chain [1] done processing
23:54:36 - cmdstanpy - INFO - Chain [1] start processing
23:54:36 - cmdstanpy - INFO - Chain [1] done processing
23:54:36 - cmdstanpy - INFO - Chain [1] start processing
23:54:36 - cmdstanpy - INFO - Chain [1] done processing
23:54:36 - cmdstanpy - INFO - Chain [1] start processing
23:54:36 - cmdstanpy - INFO - Chain [1] done processing
23:54:36 - cmdstanpy - INFO - Chain [1] start processing
23:54:36 - cmdstanpy - INFO - Chain [1] done processing
23:54:37 - cmdstanpy - INFO - Chain [1] start processing
23:54:37 - cmdstanpy - INFO - Chain [1] done processing
23:54:37 - cmdstanpy - INFO - Chain [1] start processing
23:54:37 - cmdstanpy - INFO - Chain [1] done processing
23:54:37 - cmdstanpy - INFO - Chain [1] start processing
23:54:37 - cmdstanpy - INFO - Chain [1] done processing
23:54:37 - cmdstanpy - INFO - Chain [1] start processing
23:54:37 - cmdstanpy - INFO - Chain [1] done processing
23:54:37 - cmdstanpy - INFO - Chain [1] start processing
23:54:37 - cmdstanpy - INFO - Chain [1] done processing
23:54:37 - cmdstanpy - INFO - Chain [1] start processing
23:54:38 - cmdstanpy - INFO - Chain [1] done processing
23:54:38 - cmdstanpy - INFO - Chain [1] start processing
23:54:38 - cmdstanpy - INFO - Chain [1] done processing
23:54:38 - cmdstanpy - INFO - Chain [1] start processing
23:54:38 - cmdstanpy - INFO - Chain [1] done processing
23:54:38 - cmdstanpy - INFO - Chain [1] start processing
23:54:38 - cmdstanpy - INFO - Chain [1] done processing
23:54:38 - cmdstanpy - INFO - Chain [1] start processing
23:54:38 - cmdstanpy - INFO - Chain [1] done processing
23:54:38 - cmdstanpy - INFO - Chain [1] start processing
23:54:38 - cmdstanpy - INFO - Chain [1] done processing
23:54:38 - cmdstanpy - INFO - Chain [1] start processing
23:54:39 - cmdstanpy - INFO - Chain [1] done processing
23:54:39 - cmdstanpy - INFO - Chain [1] start processing
23:54:39 - cmdstanpy - INFO - Chain [1] done processing
23:54:39 - cmdstanpy - INFO - Chain [1] start processing
23:54:39 - cmdstanpy - INFO - Chain [1] done processing
23:54:39 - cmdstanpy - INFO - Chain [1] start processing
23:54:39 - cmdstanpy - INFO - Chain [1] done processing
23:54:39 - cmdstanpy - INFO - Chain [1] start processing
23:54:40 - cmdstanpy - INFO - Chain [1] done processing
23:54:40 - cmdstanpy - INFO - Chain [1] start processing
23:54:40 - cmdstanpy - INFO - Chain [1] done processing
23:54:40 - cmdstanpy - INFO - Chain [1] start processing
23:54:40 - cmdstanpy - INFO - Chain [1] done processing
23:54:40 - cmdstanpy - INFO - Chain [1] start processing
23:54:40 - cmdstanpy - INFO - Chain [1] done processing
In [214]:
df_cv.tail()
Out[214]:
ds yhat yhat_lower yhat_upper y cutoff
111 2018-01-01 477.275960 395.544862 554.965802 336 2017-01-11
112 2018-01-01 477.275960 401.431838 551.607503 336 2017-07-10
113 2019-01-01 475.120893 395.507834 554.179964 443 2018-01-06
114 2019-01-01 475.120893 397.922470 542.395044 443 2018-07-05
115 2020-01-01 479.046162 399.273109 555.727385 397 2019-01-01
In [215]:
df_p = performance_metrics(df_cv)
df_p.head()
Out[215]:
horizon mse rmse mae mape mdape smape coverage
0 54 days 1537.324838 39.208734 27.832534 0.109518 0.058043 0.121258 0.636364
1 59 days 1529.255427 39.105696 27.254227 0.105506 0.057014 0.117096 0.636364
2 65 days 1738.022563 41.689598 31.364763 0.117560 0.098939 0.130224 0.545455
3 66 days 1524.283067 39.042068 28.114058 0.128219 0.098939 0.144690 0.545455
4 70 days 1803.177816 42.463841 30.520698 0.128677 0.098939 0.145251 0.545455
In [216]:
plot_cross_validation_metric(df_cv, metric='mae')
Out[216]:
In [217]:
plot_cross_validation_metric(df_cv, metric='mape')
Out[217]:

¶

In [218]:
plot_cross_validation_metric(df_cv, metric='rmse')
Out[218]:
In [ ]:
In [ ]:

CO2 Emissions¶

In [219]:
co2 = pd.read_csv('co2_emissions_kt_by_country.csv')
co2.head()
Out[219]:
country_code country_name year value
0 ABW Aruba 1960 11092.675
1 ABW Aruba 1961 11576.719
2 ABW Aruba 1962 12713.489
3 ABW Aruba 1963 12178.107
4 ABW Aruba 1964 11840.743
In [220]:
co2_world = co2[co2.country_name == 'World']
co2_sum = co2_world.groupby(['year'])['value'].sum().to_frame().reset_index()
In [221]:
m, b = best_fit_slope_and_intercept(pd.to_numeric(co2_sum['year']), co2_sum['value'])

regression_line = []
for x in pd.to_numeric(co2_sum['year']):
    regression_line.append((m*x)+b)
    
co2_sum['line'] = regression_line
In [222]:
# Create traces
fig = go.Figure()

fig.add_trace(go.Scatter(x = co2_sum['year'] , y=co2_sum.value, mode='lines', name='Global CO2'))
fig.add_trace(go.Scatter(x = co2_sum['year'] , y=co2_sum.line, line = dict(color='red', width=1, dash='dash'), name='Trend'))

# adjusting size of graph, legend place, and background colour
fig.update_layout(autosize=False, width=1000, height=600, margin=dict(l=50, r=50, b=100, t=100, pad=4),
    template='seaborn', title='Global CO2 Emissions', paper_bgcolor="rgb(234, 234, 242)",
    legend=dict(orientation="h", yanchor="bottom", y=1.02, xanchor="right", x=1
))
fig.update_xaxes(type='category',title='Years')
fig.update_yaxes(title='CO2 Level (kiloton)')

fig.show('notebook')
In [223]:
co2_data = co2.loc[co2['country_name']=='World']
co2_data.tail()
Out[223]:
country_code country_name year value
13647 WLD World 2015 3.299554e+07
13648 WLD World 2016 3.301856e+07
13649 WLD World 2017 3.351454e+07
13650 WLD World 2018 3.428935e+07
13651 WLD World 2019 3.434401e+07
In [224]:
co2_data['year'] = pd.to_datetime(co2_data['year'], format="%Y")

ts = co2_data[['year', 'value']].rename(columns={'year':'ds', 'value':'y'})

ts.columns=['ds','y']
co2_model = Prophet() 
co2_model.fit(ts)

/var/folders/32/vy8ww6fn78b20w6z3gzqr10h0000gn/T/ipykernel_61503/3438402208.py:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

23:54:41 - cmdstanpy - INFO - Chain [1] start processing
23:54:41 - cmdstanpy - INFO - Chain [1] done processing
Out[224]:
<prophet.forecaster.Prophet at 0x1369213c0>
In [225]:
# predict for 1 year in the furure and MS - month start is the frequency
future = co2_model.make_future_dataframe(periods = 31, freq = 'YS') 
future.tail()
Out[225]:
ds
86 2046-01-01
87 2047-01-01
88 2048-01-01
89 2049-01-01
90 2050-01-01
In [226]:
# now lets make the forecasts
forecast = co2_model.predict(future)
forecast.tail()
Out[226]:
ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper yearly yearly_lower yearly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
86 2046-01-01 5.092541e+07 4.860649e+07 5.498826e+07 4.807249e+07 5.401286e+07 8.605131e+05 8.605131e+05 8.605131e+05 8.605131e+05 8.605131e+05 8.605131e+05 0.0 0.0 0.0 5.178592e+07
87 2047-01-01 5.153086e+07 4.926466e+07 5.580849e+07 4.844704e+07 5.478895e+07 8.740582e+05 8.740582e+05 8.740582e+05 8.740582e+05 8.740582e+05 8.740582e+05 0.0 0.0 0.0 5.240492e+07
88 2048-01-01 5.213632e+07 5.001630e+07 5.673494e+07 4.884518e+07 5.557742e+07 1.047867e+06 1.047867e+06 1.047867e+06 1.047867e+06 1.047867e+06 1.047867e+06 0.0 0.0 0.0 5.318419e+07
89 2049-01-01 5.274343e+07 5.024368e+07 5.736323e+07 4.925669e+07 5.631729e+07 1.007253e+06 1.007253e+06 1.007253e+06 1.007253e+06 1.007253e+06 1.007253e+06 0.0 0.0 0.0 5.375069e+07
90 2050-01-01 5.334889e+07 5.051709e+07 5.795179e+07 4.965528e+07 5.711469e+07 8.605131e+05 8.605131e+05 8.605131e+05 8.605131e+05 8.605131e+05 8.605131e+05 0.0 0.0 0.0 5.420940e+07
In [227]:
co2_model.plot(forecast)
Out[227]:
In [228]:
df_cv = cross_validation(co2_model, initial='730 days', period='180 days', horizon = '365 days')

  0%|          | 0/114 [00:00<?, ?it/s]
23:54:42 - cmdstanpy - INFO - Chain [1] start processing
23:54:42 - cmdstanpy - INFO - Chain [1] done processing
23:54:42 - cmdstanpy - INFO - Chain [1] start processing
23:54:42 - cmdstanpy - INFO - Chain [1] done processing
23:54:42 - cmdstanpy - INFO - Chain [1] start processing
23:54:42 - cmdstanpy - INFO - Chain [1] done processing
23:54:42 - cmdstanpy - INFO - Chain [1] start processing
23:54:42 - cmdstanpy - INFO - Chain [1] done processing
23:54:42 - cmdstanpy - INFO - Chain [1] start processing
23:54:42 - cmdstanpy - INFO - Chain [1] done processing
23:54:42 - cmdstanpy - INFO - Chain [1] start processing
23:54:42 - cmdstanpy - INFO - Chain [1] done processing
23:54:42 - cmdstanpy - INFO - Chain [1] start processing
23:54:45 - cmdstanpy - INFO - Chain [1] done processing
23:54:45 - cmdstanpy - INFO - Chain [1] start processing
23:54:47 - cmdstanpy - INFO - Chain [1] done processing
23:54:47 - cmdstanpy - INFO - Chain [1] start processing
23:54:50 - cmdstanpy - INFO - Chain [1] done processing
23:54:50 - cmdstanpy - INFO - Chain [1] start processing
23:54:52 - cmdstanpy - INFO - Chain [1] done processing
23:54:52 - cmdstanpy - INFO - Chain [1] start processing
23:54:52 - cmdstanpy - INFO - Chain [1] done processing
23:54:52 - cmdstanpy - INFO - Chain [1] start processing
23:54:52 - cmdstanpy - INFO - Chain [1] done processing
23:54:52 - cmdstanpy - INFO - Chain [1] start processing
23:54:52 - cmdstanpy - INFO - Chain [1] done processing
23:54:52 - cmdstanpy - INFO - Chain [1] start processing
23:54:53 - cmdstanpy - INFO - Chain [1] done processing
23:54:53 - cmdstanpy - INFO - Chain [1] start processing
23:54:56 - cmdstanpy - INFO - Chain [1] done processing
23:54:56 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:54:59 - cmdstanpy - INFO - Chain [1] start processing
23:54:59 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:00 - cmdstanpy - INFO - Chain [1] done processing
23:55:00 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:01 - cmdstanpy - INFO - Chain [1] done processing
23:55:01 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:02 - cmdstanpy - INFO - Chain [1] done processing
23:55:02 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:03 - cmdstanpy - INFO - Chain [1] done processing
23:55:03 - cmdstanpy - INFO - Chain [1] start processing
23:55:04 - cmdstanpy - INFO - Chain [1] done processing
23:55:04 - cmdstanpy - INFO - Chain [1] start processing
23:55:04 - cmdstanpy - INFO - Chain [1] done processing
23:55:04 - cmdstanpy - INFO - Chain [1] start processing
23:55:04 - cmdstanpy - INFO - Chain [1] done processing
23:55:04 - cmdstanpy - INFO - Chain [1] start processing
23:55:04 - cmdstanpy - INFO - Chain [1] done processing
23:55:04 - cmdstanpy - INFO - Chain [1] start processing
23:55:04 - cmdstanpy - INFO - Chain [1] done processing
23:55:04 - cmdstanpy - INFO - Chain [1] start processing
23:55:04 - cmdstanpy - INFO - Chain [1] done processing
23:55:04 - cmdstanpy - INFO - Chain [1] start processing
23:55:04 - cmdstanpy - INFO - Chain [1] done processing
23:55:04 - cmdstanpy - INFO - Chain [1] start processing
23:55:05 - cmdstanpy - INFO - Chain [1] done processing
23:55:05 - cmdstanpy - INFO - Chain [1] start processing
23:55:05 - cmdstanpy - INFO - Chain [1] done processing
23:55:05 - cmdstanpy - INFO - Chain [1] start processing
23:55:05 - cmdstanpy - INFO - Chain [1] done processing
23:55:05 - cmdstanpy - INFO - Chain [1] start processing
23:55:05 - cmdstanpy - INFO - Chain [1] done processing
23:55:05 - cmdstanpy - INFO - Chain [1] start processing
23:55:05 - cmdstanpy - INFO - Chain [1] done processing
23:55:05 - cmdstanpy - INFO - Chain [1] start processing
23:55:05 - cmdstanpy - INFO - Chain [1] done processing
23:55:05 - cmdstanpy - INFO - Chain [1] start processing
23:55:06 - cmdstanpy - INFO - Chain [1] done processing
23:55:06 - cmdstanpy - INFO - Chain [1] start processing
23:55:06 - cmdstanpy - INFO - Chain [1] done processing
23:55:06 - cmdstanpy - INFO - Chain [1] start processing
23:55:06 - cmdstanpy - INFO - Chain [1] done processing
23:55:06 - cmdstanpy - INFO - Chain [1] start processing
23:55:06 - cmdstanpy - INFO - Chain [1] done processing
23:55:06 - cmdstanpy - INFO - Chain [1] start processing
23:55:06 - cmdstanpy - INFO - Chain [1] done processing
23:55:06 - cmdstanpy - INFO - Chain [1] start processing
23:55:06 - cmdstanpy - INFO - Chain [1] done processing
23:55:06 - cmdstanpy - INFO - Chain [1] start processing
23:55:06 - cmdstanpy - INFO - Chain [1] done processing
23:55:06 - cmdstanpy - INFO - Chain [1] start processing
23:55:07 - cmdstanpy - INFO - Chain [1] done processing
23:55:07 - cmdstanpy - INFO - Chain [1] start processing
23:55:07 - cmdstanpy - INFO - Chain [1] done processing
23:55:07 - cmdstanpy - INFO - Chain [1] start processing
23:55:07 - cmdstanpy - INFO - Chain [1] done processing
00:09:03 - cmdstanpy - INFO - Chain [1] start processing
00:09:03 - cmdstanpy - INFO - Chain [1] done processing
00:09:03 - cmdstanpy - INFO - Chain [1] start processing
00:09:04 - cmdstanpy - INFO - Chain [1] done processing
00:09:04 - cmdstanpy - INFO - Chain [1] start processing
00:09:04 - cmdstanpy - INFO - Chain [1] done processing
00:09:04 - cmdstanpy - INFO - Chain [1] start processing
00:09:04 - cmdstanpy - INFO - Chain [1] done processing
00:09:04 - cmdstanpy - INFO - Chain [1] start processing
00:09:04 - cmdstanpy - INFO - Chain [1] done processing
00:09:04 - cmdstanpy - INFO - Chain [1] start processing
00:09:04 - cmdstanpy - INFO - Chain [1] done processing
00:09:04 - cmdstanpy - INFO - Chain [1] start processing
00:09:05 - cmdstanpy - INFO - Chain [1] done processing
00:09:05 - cmdstanpy - INFO - Chain [1] start processing
00:09:05 - cmdstanpy - INFO - Chain [1] done processing
00:09:05 - cmdstanpy - INFO - Chain [1] start processing
00:09:05 - cmdstanpy - INFO - Chain [1] done processing
00:09:05 - cmdstanpy - INFO - Chain [1] start processing
00:09:05 - cmdstanpy - INFO - Chain [1] done processing
00:09:05 - cmdstanpy - INFO - Chain [1] start processing
00:09:06 - cmdstanpy - INFO - Chain [1] done processing
00:09:06 - cmdstanpy - INFO - Chain [1] start processing
00:09:06 - cmdstanpy - INFO - Chain [1] done processing
00:09:06 - cmdstanpy - INFO - Chain [1] start processing
00:09:06 - cmdstanpy - INFO - Chain [1] done processing
00:09:06 - cmdstanpy - INFO - Chain [1] start processing
00:09:06 - cmdstanpy - INFO - Chain [1] done processing
00:09:06 - cmdstanpy - INFO - Chain [1] start processing
00:09:06 - cmdstanpy - INFO - Chain [1] done processing
00:09:06 - cmdstanpy - INFO - Chain [1] start processing
00:09:06 - cmdstanpy - INFO - Chain [1] done processing
00:09:06 - cmdstanpy - INFO - Chain [1] start processing
00:09:07 - cmdstanpy - INFO - Chain [1] done processing
00:09:07 - cmdstanpy - INFO - Chain [1] start processing
00:09:07 - cmdstanpy - INFO - Chain [1] done processing
00:09:07 - cmdstanpy - INFO - Chain [1] start processing
00:09:07 - cmdstanpy - INFO - Chain [1] done processing
00:09:07 - cmdstanpy - INFO - Chain [1] start processing
00:09:07 - cmdstanpy - INFO - Chain [1] done processing
00:09:07 - cmdstanpy - INFO - Chain [1] start processing
00:09:07 - cmdstanpy - INFO - Chain [1] done processing
00:09:07 - cmdstanpy - INFO - Chain [1] start processing
00:09:07 - cmdstanpy - INFO - Chain [1] done processing
00:09:07 - cmdstanpy - INFO - Chain [1] start processing
00:09:08 - cmdstanpy - INFO - Chain [1] done processing
00:09:08 - cmdstanpy - INFO - Chain [1] start processing
00:09:08 - cmdstanpy - INFO - Chain [1] done processing
00:09:08 - cmdstanpy - INFO - Chain [1] start processing
00:09:08 - cmdstanpy - INFO - Chain [1] done processing
00:09:08 - cmdstanpy - INFO - Chain [1] start processing
00:09:08 - cmdstanpy - INFO - Chain [1] done processing
00:09:08 - cmdstanpy - INFO - Chain [1] start processing
00:09:08 - cmdstanpy - INFO - Chain [1] done processing
00:09:08 - cmdstanpy - INFO - Chain [1] start processing
00:09:09 - cmdstanpy - INFO - Chain [1] done processing
00:09:09 - cmdstanpy - INFO - Chain [1] start processing
00:09:09 - cmdstanpy - INFO - Chain [1] done processing
00:09:09 - cmdstanpy - INFO - Chain [1] start processing
00:09:09 - cmdstanpy - INFO - Chain [1] done processing
In [229]:
df_cv.head()
Out[229]:
ds yhat yhat_lower yhat_upper y cutoff
0 1963-01-01 1.003124e+07 1.003124e+07 1.003124e+07 1.027450e+07 1962-04-25
1 1963-01-01 1.003124e+07 1.003124e+07 1.003124e+07 1.027450e+07 1962-10-22
2 1964-01-01 1.106801e+07 1.106327e+07 1.107313e+07 1.082025e+07 1963-04-20
3 1964-01-01 1.106801e+07 1.106301e+07 1.107294e+07 1.082025e+07 1963-10-17
4 1965-01-01 1.099053e+07 1.093655e+07 1.104658e+07 1.140258e+07 1964-04-14
In [230]:
df_p = performance_metrics(df_cv)
df_p.head()
Out[230]:
horizon mse rmse mae mape mdape smape coverage
0 54 days 2.414025e+12 1.553713e+06 1.300441e+06 0.061327 0.040427 0.059135 0.272727
1 59 days 2.408448e+12 1.551918e+06 1.282817e+06 0.060460 0.040427 0.058275 0.272727
2 64 days 2.380311e+12 1.542826e+06 1.252229e+06 0.058995 0.040427 0.056778 0.363636
3 70 days 2.382926e+12 1.543673e+06 1.255385e+06 0.059079 0.040427 0.056863 0.363636
4 71 days 2.355546e+12 1.534779e+06 1.222926e+06 0.058791 0.040427 0.056568 0.272727
In [231]:
plot_cross_validation_metric(df_cv, metric='rmse')
Out[231]:
In [232]:
plot_cross_validation_metric(df_cv, metric='mae')
Out[232]:
In [233]:
plot_cross_validation_metric(df_cv, metric='mape')
Out[233]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:

Sea level change¶

In [234]:
sea_pd = pd.read_csv('sea_levels_2015.csv')
sea_pd['Time'] = pd.to_datetime(sea_pd['Time'], format="%Y/%m/%d")
sea_pd = sea_pd.loc[(sea_pd['Time'] >= '1960-01-01') & (sea_pd['Time'] <= '2019-12-31')]
sea_pd.head()
Out[234]:
Time GMSL GMSL uncertainty
960 1960-01-15 -48.9 6.6
961 1960-02-15 -44.9 6.6
962 1960-03-15 -45.1 6.6
963 1960-04-15 -43.1 6.6
964 1960-05-15 -42.5 6.6
In [235]:
# Create traces
fig = go.Figure()

fig.add_trace(go.Scatter(x = sea_pd['Time'] , y=sea_pd.GMSL, mode='lines', name='GMSL'))
#fig.add_trace(go.Scatter(x = sea_pd['Time'] , y=sea_pd.line, line = dict(color='red', width=1, dash='dash'), name='Trend'))

# adjusting size of graph, legend place, and background colour
fig.update_layout(autosize=False, width=1000, height=600, margin=dict(l=50,r=50, b=100,t=100,pad=4),
    template='seaborn', title='Global Mean Sea Level changes', paper_bgcolor="rgb(234, 234, 242)",
    legend=dict(orientation="h", yanchor="bottom",y=1.02,xanchor="right", x=1))
fig.update_xaxes(type='category',title='Years')
fig.update_yaxes(title='Global Mean Sea Level (mm)')

fig.show('notebook')
In [236]:
ts=sea_pd.groupby(["Time"])["GMSL"].sum()
In [237]:
ts = sea_pd.rename(columns={'Time':'ds', 'GMSL':'y','GMSL uncertainty':'yhat'})

ts.columns=['ds','y','yhat']
model1 = Prophet( yearly_seasonality=True) 
model1.fit(ts)

00:09:10 - cmdstanpy - INFO - Chain [1] start processing
00:09:10 - cmdstanpy - INFO - Chain [1] done processing
Out[237]:
<prophet.forecaster.Prophet at 0x15388f910>
In [238]:
# predict for 1 year in the furure and MS - month start is the frequency
future = model1.make_future_dataframe(periods = 444, freq = 'MS') 
future.tail()
Out[238]:
ds
1087 2050-08-01
1088 2050-09-01
1089 2050-10-01
1090 2050-11-01
1091 2050-12-01
In [239]:
# now lets make the forecasts
forecast = model1.predict(future)
forecast.tail()
Out[239]:
ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper yearly yearly_lower yearly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
1087 2050-08-01 219.629618 195.112741 250.663533 191.263928 245.210170 4.880047 4.880047 4.880047 4.880047 4.880047 4.880047 0.0 0.0 0.0 224.509665
1088 2050-09-01 219.966233 202.783910 257.425592 191.521937 245.606808 11.812785 11.812785 11.812785 11.812785 11.812785 11.812785 0.0 0.0 0.0 231.779018
1089 2050-10-01 220.291990 194.371609 249.744197 191.774540 245.998627 3.272461 3.272461 3.272461 3.272461 3.272461 3.272461 0.0 0.0 0.0 223.564451
1090 2050-11-01 220.628604 187.159152 243.578743 192.048011 246.407347 -3.849131 -3.849131 -3.849131 -3.849131 -3.849131 -3.849131 0.0 0.0 0.0 216.779473
1091 2050-12-01 220.954361 180.715191 236.260323 192.316411 246.835055 -11.293515 -11.293515 -11.293515 -11.293515 -11.293515 -11.293515 0.0 0.0 0.0 209.660845
In [240]:
forecast.head()
Out[240]:
ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper yearly yearly_lower yearly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
0 1960-01-15 -47.612698 -55.338371 -41.237153 -47.612698 -47.612698 -0.617279 -0.617279 -0.617279 -0.617279 -0.617279 -0.617279 0.0 0.0 0.0 -48.229977
1 1960-02-15 -47.545456 -55.462295 -40.950030 -47.545456 -47.545456 -0.290799 -0.290799 -0.290799 -0.290799 -0.290799 -0.290799 0.0 0.0 0.0 -47.836254
2 1960-03-15 -47.482552 -55.074663 -40.554056 -47.482552 -47.482552 -0.371233 -0.371233 -0.371233 -0.371233 -0.371233 -0.371233 0.0 0.0 0.0 -47.853785
3 1960-04-15 -47.415310 -54.904377 -40.519979 -47.415310 -47.415310 0.043208 0.043208 0.043208 0.043208 0.043208 0.043208 0.0 0.0 0.0 -47.372102
4 1960-05-15 -47.350237 -54.536069 -39.356471 -47.350237 -47.350237 0.408130 0.408130 0.408130 0.408130 0.408130 0.408130 0.0 0.0 0.0 -46.942107
In [241]:
model1.plot(forecast)
Out[241]:
In [244]:
df_cv = cross_validation(model1, initial='730 days', period='180 days', horizon = '365 days')

  0%|          | 0/104 [00:00<?, ?it/s]
00:09:11 - cmdstanpy - INFO - Chain [1] start processing
00:09:17 - cmdstanpy - INFO - Chain [1] done processing
00:09:17 - cmdstanpy - INFO - Chain [1] start processing
00:09:17 - cmdstanpy - INFO - Chain [1] done processing
00:09:17 - cmdstanpy - INFO - Chain [1] start processing
00:09:17 - cmdstanpy - INFO - Chain [1] done processing
00:09:17 - cmdstanpy - INFO - Chain [1] start processing
00:09:17 - cmdstanpy - INFO - Chain [1] done processing
00:09:17 - cmdstanpy - INFO - Chain [1] start processing
00:09:18 - cmdstanpy - INFO - Chain [1] done processing
00:09:18 - cmdstanpy - INFO - Chain [1] start processing
00:09:18 - cmdstanpy - INFO - Chain [1] done processing
00:09:18 - cmdstanpy - INFO - Chain [1] start processing
00:09:18 - cmdstanpy - INFO - Chain [1] done processing
00:09:18 - cmdstanpy - INFO - Chain [1] start processing
00:09:18 - cmdstanpy - INFO - Chain [1] done processing
00:09:18 - cmdstanpy - INFO - Chain [1] start processing
00:09:18 - cmdstanpy - INFO - Chain [1] done processing
00:09:18 - cmdstanpy - INFO - Chain [1] start processing
00:09:18 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:19 - cmdstanpy - INFO - Chain [1] done processing
00:09:19 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:20 - cmdstanpy - INFO - Chain [1] done processing
00:09:20 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:21 - cmdstanpy - INFO - Chain [1] done processing
00:09:21 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:22 - cmdstanpy - INFO - Chain [1] start processing
00:09:22 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:23 - cmdstanpy - INFO - Chain [1] done processing
00:09:23 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:24 - cmdstanpy - INFO - Chain [1] start processing
00:09:24 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
00:09:25 - cmdstanpy - INFO - Chain [1] start processing
00:09:25 - cmdstanpy - INFO - Chain [1] done processing
In [245]:
df_cv.head()
Out[245]:
ds yhat yhat_lower yhat_upper y cutoff
0 1962-03-15 -40.634492 -40.803489 -40.435703 -44.1 1962-03-13
1 1962-04-15 -45.908039 -46.420057 -45.280025 -44.6 1962-03-13
2 1962-05-15 -51.670550 -52.608106 -50.461577 -45.4 1962-03-13
3 1962-06-15 -55.682967 -57.239274 -53.814497 -45.1 1962-03-13
4 1962-07-15 -56.358464 -58.536608 -53.697625 -46.6 1962-03-13
In [246]:
df_p = performance_metrics(df_cv)
df_p.head()
Out[246]:
horizon mse rmse mae mdape smape coverage
0 37 days 47.608001 6.899855 5.757335 0.226597 0.476291 0.625000
1 38 days 47.978280 6.926636 5.771935 0.226597 0.486250 0.629032
2 39 days 48.085070 6.934340 5.766684 0.220859 0.477309 0.629032
3 40 days 48.466069 6.961758 5.823585 0.230476 0.482648 0.618280
4 41 days 48.460874 6.961384 5.809885 0.230476 0.479929 0.616935
In [247]:
df_p.tail()
Out[247]:
horizon mse rmse mae mdape smape coverage
324 361 days 65.271436 8.079074 6.653391 0.240987 0.518110 0.572581
325 362 days 64.163743 8.010227 6.583705 0.228181 0.512906 0.569892
326 363 days 63.713564 7.982078 6.585724 0.240987 0.529785 0.561828
327 364 days 63.880574 7.992532 6.590234 0.228181 0.529507 0.560484
328 365 days 68.449301 8.273409 6.839334 0.243107 0.539219 0.546371
In [248]:
plot_cross_validation_metric(df_cv, metric='rmse')
Out[248]:
In [249]:
plot_cross_validation_metric(df_cv, metric='mae')
Out[249]:
In [ ]:
In [ ]: